In May 2026 Gartner reported that sales organisations providing sellers with AI-enabled next best actions were 2.6 times more likely to achieve commercial growth. The figure is already circulating in vendor material as though it were a return on investment.
What a multiple like this can and cannot support
It is an association between two characteristics of surveyed organisations. It does not establish that the next best actions caused the growth, and the most likely confound is easy to name: firms mature enough to generate a useful next best action are firms that already have clean records, defined stages, agreed definitions and someone accountable for the pipeline.
Those same properties independently predict commercial growth. A firm that buys the recommendation layer without the underlying discipline has bought the visible part of the correlation.
None of that makes the finding useless. It makes it a finding about a bundle rather than about a feature, and the bundle is the thing worth acquiring.
The prerequisite nobody sells
A next best action is a recommendation computed from a record. If the record is a forecasting artefact filled in under duress, the recommendation is computed from fiction and delivered with the confidence of software, which is worse than no recommendation because it is harder to argue with.
The sequence that follows from this is unglamorous. Define the stages so two people would agree on which one a deal is in. Capture what the buyer said rather than what stage they are at. Automate the capture, because it is production work and a seller doing it is a judgment seat doing data entry. Then add recommendation on top of a record that can support it.
What we take from it
We read the 2.6x as evidence that commercial maturity pays, which we already believed, rather than as evidence that a particular feature pays, which it cannot establish.
The general habit is worth more than this particular number. When a vendor-adjacent statistic quantifies the benefit of a thing that vendor sells, the first question is what else is true of the firms in the numerator. It is usually the answer to your actual question.